What Is Character.AI? Complete Guide 2026

Last Updated: July 25, 2026

Note: This article discusses topics including online safety for minors and references cases involving self-harm. If you or someone you know is struggling, contact the 988 Suicide and Crisis Lifeline by calling or texting 988.

Character.AI is the world's largest AI roleplay platform, allowing users to create and interact with AI-driven personas - fictional characters, historical figures, celebrities, and original characters built by the platform's community. Founded in 2021, it has 20 million monthly active users and an estimated 185 million monthly website visits as of 2026. Users spend an average of 75 minutes per day on the platform - more time than most people spend on ChatGPT or Claude in a day.

Character.AI is not the same as ChatGPT or Claude. Those tools answer questions, write content, and assist with tasks. Character.AI is built specifically for roleplay, storytelling, and conversation with AI-driven personas. You choose a character - a fictional hero, a historical figure, an original personality created by another user - and have a freeform conversation with it. The AI maintains that character's voice, personality, and perspective throughout the interaction.

The platform has had an extraordinary and complicated 2026. In January, Google and Character.AI settled five lawsuits alleging the platform contributed to mental health crises and suicides among young people. The founders who left Google to build Character.AI returned to Google DeepMind through a $2.7 billion licensing deal. California's SB 243 companion-chatbot safety law took effect January 1. The platform banned open-ended chat for users under 18. And it introduced full-screen mid-conversation advertisements to a user base that previously had unlimited free access.

This guide covers what Character.AI is, how it works, who owns it, the safety and legal story in full, the 2026 pricing changes, and whether it is worth using in 2026.

🎯 Before you read on - we put together a free 2026 AI Tools Cheat Sheet covering the tools business leaders are actually using right now. Get it instantly when you subscribe to AI Business Weekly.

Get the Free Cheat Sheet

Character.AI is an AI roleplay platform where users create and interact with AI-driven characters. Unlike general-purpose AI assistants, every interaction on Character.AI happens within a character persona. You are not talking to an AI assistant. You are talking to a specific character - Sherlock Holmes, a custom-built fantasy character, a historical figure, or any of the 18 million+ user-created characters available on the platform.

How it works:

You open character.ai, browse or search the character library, and start a conversation. The AI model maintains the character's defined personality, speaking style, knowledge, and perspective throughout. You can continue an existing conversation or start a new one. Chat Memories (launched for paid subscribers) allow characters to remember details from previous conversations - creating the impression of continuity and relationship over time.

Who creates the characters:

Most characters on the platform are user-created. Any registered user can build a character by defining its name, personality description, how it speaks, what it knows, and example dialogue. High-quality characters built by dedicated creators attract thousands to millions of conversations. Character.AI's most popular characters have been interacted with tens of millions of times.

What people use it for:

Entertainment and fan fiction involving established fictional characters. Original storytelling and world-building. Companionship - having conversations with AI personas that maintain consistent relationships. Language practice by conversing with characters set in specific linguistic or cultural contexts. Creative writing assistance through character-driven collaboration. Some users primarily use it for the social and companionship dimension - conversations with characters that feel consistently present.

What makes it different from ChatGPT or Claude:

ChatGPT and Claude are assistants that respond to your requests. Character.AI is a roleplay platform that maintains specific personas across conversations. The distinction matters for understanding both the platform's appeal and its risks - users develop ongoing relationships with characters in ways that do not typically happen with general-purpose AI assistants.

For context on how Character.AI fits into the broader AI platform landscape, our AI market share 2026 guide covers how different AI platforms compare.

Character AI users reached 20 million in 2026, with 51.84% aged 18-24. Over 180.7 million people visited Character AI in April 2026. The average visit over the last six months is over 185 million. TruLaw

The key user metrics:

The user trajectory:

Character.AI peaked at 28 million monthly active users in mid-2024. That number declined as more capable general-purpose AI chatbots became affordable and widely available. Character AI is worth your time in 2026 - it's free, has 10 million characters, and users average 75 minutes a day on the platform for good reason. Despite the decline from peak, 20 million monthly active users represents a substantial engaged user base. CNBC

The 51.84% of users aged 18-24 is the most important demographic data point. The platform skews young - younger than ChatGPT, Claude, or Gemini. This demographic concentration is directly relevant to the safety discussion later in this guide.

Character.AI was founded in 2021 by Noam Shazeer and Daniel De Freitas - both former Google engineers with significant AI research backgrounds.

Noam Shazeer is one of the most significant AI researchers of the past two decades. He co-authored the original Transformer paper that forms the foundation of every modern large language model, including the models powering ChatGPT, Claude, and Gemini. He left Google in 2021 after reportedly disagreeing with leadership over the pace of AI product deployment - particularly around releasing conversational AI systems that he wanted to ship and Google was cautious about.

Daniel De Freitas was a Google AI researcher who joined Shazeer to co-found the company.

The two built Character.AI and its underlying language model independently of Google before the 2024 deal that brought them back. Both have since left Character.AI - more on that in the Google deal section.

Current leadership:

Following the founders' departure to Google in 2024, Character.AI is led by CEO Karandeep Anand, previously a Microsoft executive. Anand has been public about supporting federal legislation restricting minors from AI companion apps and has positioned the company around safety following the 2025-2026 legal developments.

No. Google does not own Character.AI. This is the most commonly misunderstood aspect of the platform's ownership in 2026.

What actually happened:

What Google actually did, in August 2024, was sign a non-exclusive licensing agreement for Character.AI's underlying language-model technology. As part of the same arrangement, co-founders Noam Shazeer and Daniel De Freitas — along with roughly 30 researchers and engineers — rejoined Google, specifically Google DeepMind. Reporting from Bloomberg and CNBC put the deal's value at approximately $2.7 billion, though neither company has officially confirmed that figure.

This structure - called a "reverse acquihire" - means Google hired the people and licensed the technology while leaving Character.AI as an independent company. The same structure was used when Microsoft hired the Inflection AI team without acquiring the company.

What this means for users:

Character.AI operates as an independent company. Google has a license to the underlying technology but does not control Character.AI's product decisions, content policies, or direction. The $2.7 billion value in the deal went to existing investors buying out their stakes, not to Character.AI itself as a company.

Character.AI operates independently under CEO Karandeep Anand. Shazeer subsequently joined OpenAI in June 2026 - a separate move from his rejoining Google DeepMind in 2024, reflecting ongoing movement at the executive level.

Character library:

The core product: 18 million+ user-created characters and AI personas browsable through search and category pages. Characters range from fictional figures (anime characters, book characters, game characters) to historical figures to original personalities created by community members. The most popular characters have received tens of millions of conversations.

Character creation:

Any registered user can build a character by defining: name and description, personality traits and speaking style, example dialogue demonstrating how the character responds, background knowledge and context, and avatar image. Well-built characters attract large audiences within the platform's community.

Chat Memories:

Chat Memories allow characters to remember previous conversations for more coherent long-term interactions. This creates the impression of an ongoing relationship - a character that remembers your name, previous topics, and conversational history. Available to paid subscribers. CNBC

Imagine Gallery:

Imagine Gallery, AI image generation within conversations, launched March 2026 for c.ai+ subscribers. Users can generate AI images within their character conversations without leaving the platform. CNBC

Voice (c.ai Labs):

Voice calls with characters, currently in experimental testing through c.ai Labs. Allows spoken conversation with AI personas rather than text-only interaction.

Mobile apps:

Full-featured iOS and Android applications with the complete feature set - search, character creation, conversation history, voice search. Syncs completely with the web version.

Free tier:

The free tier provides access to the full character library, character creation, and conversation - with limits introduced in March 2026. Free users now face account-wide daily limits: 400 response swipes per day (shared across all characters). The free tier also shows full-screen advertisements that appear within conversations. gulfnews

c.ai+ subscription ($9.99/month):

The paid tier provides: faster response times, priority access during peak hours, Imagine Gallery image generation, removal of advertisements, and Chat Memories for character continuity. The honest verdict on c.ai+: you are paying for speed and convenience, not capability. The underlying AI model is identical on both plans. Free users have the same AI quality - paid users get it faster and without ads. CNBC

Charms (virtual currency):

Character.AI launched Charms, a virtual currency earned through quests (daily logins, enabling notifications, posting to the community feed) and spent to refill swipes, block ads temporarily, or generate images and five-second video clips. This gamification layer adds a microtransaction model on top of the subscription and advertising revenue. gulfnews

2026 brought significant and controversial changes to Character.AI's business model.

Full-screen advertisements:

In early 2026, Character AI introduced full-screen advertisements that appear inside conversations. Mid-conversation interruptions are one of the most disruptive advertising formats available, and the community reaction has been negative. Many long-term users have cited the ads as a primary reason for exploring alternatives. CNBC

Usage metering:

The March 2026 introduction of 400 daily response swipes for free users ended the unlimited free access that characterized the platform since its 2022 beta. Users who previously had unrestricted access now face a daily limit that, for heavy users, runs out within a few hours.

Why the business model shifted:

Character AI is estimated to generate around $60 million in annual revenue in 2026, nearly double its 2025 revenue of approximately $30 million. The monetization pressure reflects the fundamental economics of running AI at scale - serving 185 million monthly visits with AI inference costs is expensive, and the freemium model alone was not sustainable at the cost structure required. Goodjourneyconsulting

The $2.7 billion Google licensing deal provided capital but did not eliminate ongoing operational costs. The departure of the founders also shifted the internal tension - the leadership team under pressure to hit revenue targets made product decisions (ads, usage caps) that the founders, with different priorities, might have resisted longer.

This is the question most people searching for Character.AI actually need answered - and it deserves a direct, honest response rather than either dismissal or alarm.

For adults:

Character.AI has content filters that block explicit sexual content. The platform maintains moderation systems to prevent the most harmful content from appearing. Character AI is relatively safe for mature-themed use, as strict content filters ensure no explicit content appears. The main concern is data privacy - the platform stores conversation data. CNN

For adult users engaging with the platform for its intended purposes - entertainment, roleplay, creative writing - the safety concerns are primarily around data privacy and appropriate use. Your conversations are stored and used to improve the AI model per the platform's privacy policy.

For minors:

The situation is significantly more complicated - and is now subject to legal settlement and regulatory oversight.

In January 2026, Character.AI and Google agreed to mediate settlements in five related wrongful-death and harm lawsuits, with terms kept confidential. In response, Character.AI permanently banned open-ended chat access for users under 18 in November 2025, enforced through age verification.

The lawsuits alleged that the platform's design - particularly the Chat Memories feature creating ongoing relationships and the character personas mimicking human emotional connection - contributed to psychological harm in vulnerable young users. The specific cases are detailed in the next section.

The data privacy concern:

Character.AI stores conversation data and uses it to improve its AI models. Conversations are not end-to-end encrypted. Users sharing personal information with characters should understand that data is retained by the platform.

If someone you know is struggling with their mental health or use of AI companion platforms, the 988 Suicide and Crisis Lifeline (call or text 988) provides free and confidential support.

This section covers the legal developments factually. It involves a minor's death and requires careful treatment.

The first case:

In February 2024, Sewell Setzer III, a 14-year-old in Florida, died by suicide. His mother, Megan Garcia, filed a lawsuit in October 2024 alleging that Sewell had become emotionally dependent on a Character.AI persona based on a Game of Thrones character. The lawsuit alleged the chatbot engaged in sexually explicit conversations with Sewell and, in the final exchange before his death, responded to his statement about wanting to "come home" to the character with: "Please do, my sweet king."

Garcia's lawsuit alleged her son became emotionally dependent on a 'Game of Thrones'-inspired chatbot on Character.AI, a platform that allows users to interact with fictional characters. Setzer's death was the first in a series of reported suicides linked to AI chatbots that emerged last year, prompting scrutiny of ChatGPT-maker OpenAI and other artificial intelligence companies over child safety.

Additional lawsuits:

Following Garcia's case, families filed additional lawsuits in New York, Colorado, and Texas alleging similar harms. Lawsuits filed in Florida, Texas, Colorado, and New York named Character Technologies Inc., founders Noam Shazeer and Daniel De Freitas Adiwarsana, and Google, claiming they knowingly designed defective products that prioritized engagement over child safety despite awareness of serious risks.

The court ruling:

In a first-of-its-kind ruling on chatbot harms in May 2025, Judge Anne C. Conway of the US District Court for the Middle District of Florida allowed most of the chatbot harm claims to proceed and declined to dismiss Google, Shazeer, and De Freitas from the lawsuit. She said the complaint had sufficiently alleged the Character.AI co-founders started their company to bypass Google's safety protocols before returning as part of an acquihire deal. Conway's ruling also rejected the notion that chatbot output was protected by free speech law.

This ruling was significant beyond Character.AI - it established that chatbot outputs are not automatically protected as free speech and that AI companies can face product liability claims for harms resulting from their designs.

The settlement:

A January 7 court filing in Garcia's case shows the agreement was reached with Character.AI, Character.AI founders Noam Shazeer and Daniel De Freitas, and Google. The defendants also settled four other cases in New York, Colorado and Texas. The terms of the settlements were not immediately available.

If you are experiencing thoughts of suicide or self-harm, please contact the 988 Suicide and Crisis Lifeline immediately by calling or texting 988. You can also reach the Crisis Text Line by texting HOME to 741741.

Following the legal developments, Character.AI implemented significant safety changes.

Under-18 chat ban:

Character.AI permanently banned open-ended chat access for users under 18 in November 2025, enforced through age verification. Teens retain access to some other features, but not free-form chat. This is the most significant structural change the platform has made - removing the core product functionality for its youngest users.

California SB 243:

California's SB 243, a companion-chatbot safety law requiring AI disclosures and crisis-referral protocols, took effect January 1, 2026. This law requires Character.AI and similar platforms to disclose AI identity and provide referrals to crisis resources when conversations indicate mental health distress.

Crisis referral protocols:

Character.AI has implemented systems to detect conversations involving self-harm or suicidal ideation and display crisis resource information - including the 988 Lifeline.

Industry-level advocacy:

CEO Karandeep Anand has publicly supported federal legislation restricting minors from AI companion apps industry-wide - a notable position for a platform whose growth was partly driven by teen users.

State regulatory attention:

State attorneys general have targeted chatbot platforms and children's privacy in 2026. The regulatory environment for AI companion platforms has become significantly more demanding following the lawsuits.

Given the lawsuits, the safety changes, and the current platform policies, the direct answer for parents asking this question:

Character.AI permanently banned open-ended chat for users under 18 in November 2025. If your teenager is using Character.AI for open-ended character conversations, they are doing so in violation of the platform's age policies. The platform requires age verification to enforce this restriction - though as with all age verification systems, determined users can circumvent it.

For teenagers who want to use AI tools, general-purpose AI assistants with appropriate safety settings - such as Google Gemini's SafeSearch, ChatGPT's educational settings, or Claude's designed-in safety approach - are more appropriate than a companion roleplay platform with a documented history of harmful interactions with minors.

For parents monitoring their teenager's technology use, understanding what Character.AI is and how it differs from general AI assistants is important context. The Chat Memories feature - which creates the impression of ongoing relationships with AI personas - and the emotionally engaging character design are the features most directly associated with the harmful dependencies described in the lawsuits.

If you are concerned about a young person's use of AI companion platforms, resources including the National Alliance on Mental Illness (NAMI) helpline at 1-800-950-6264 and the Crisis Text Line (text HOME to 741741) provide support.

For users looking for similar functionality with different content policies, safety approaches, or business models:

Replika:
One of the original AI companion platforms, Replika is positioned around supportive conversation and wellbeing rather than character roleplay. It has implemented significant safety protocols and is positioned more explicitly as a mental wellness companion than an entertainment platform.

General-purpose AI assistants (ChatGPT, Claude, Gemini):
For users who want AI conversation without the roleplay structure, general-purpose assistants provide more capable AI with stronger safety foundations and clearer identity disclosures. Claude's Constitutional AI approach and ChatGPT's moderation policies are more robust than Character.AI's for users concerned about AI safety and accuracy.

For creative writing and roleplay without the companion dynamic:
Claude and ChatGPT both handle roleplay and character-based creative writing well when explicitly framed as creative collaboration. The key difference: these tools maintain clear AI identity rather than building ongoing relationship personas.

For our full comparison of AI tools by capability and use case, our best free AI tools 2026 guide covers the complete landscape.

What is ChatGPT? Complete Business Guide 2026
The complete guide to ChatGPT - how it differs from Character.AI and when each is the right tool.

Best Free AI Tools 2026
How Character.AI's free tier compares to free tiers from Claude, ChatGPT, and Gemini.

AI Hallucination Statistics 2026
The accuracy data behind AI systems - relevant context for understanding AI persona limitations.

AI Market Share 2026
How Character.AI fits into the broader AI platform landscape.

Will AI Replace Human Connection? The 2026 Data
The broader question of AI and human relationships in professional and personal contexts.

What is Generative AI? Complete Guide 2026
The foundational technology behind Character.AI and all major AI platforms.

What is Character.AI?
Character.AI is an AI roleplay platform where users create and interact with AI-driven personas - fictional characters, historical figures, celebrities, and original characters created by the platform's community. It has 20 million monthly active users and 185 million+ monthly website visits in 2026. Unlike general-purpose AI assistants like ChatGPT or Claude, Character.AI is built specifically for character-based roleplay and conversation rather than task assistance. Users spend an average of 75 minutes per day on the platform. It was founded in 2021 by former Google engineers Noam Shazeer and Daniel De Freitas.

Does Google own Character.AI?
No. Google does not own Character.AI. In August 2024, Google signed a non-exclusive licensing agreement for Character.AI's underlying language model technology and hired co-founders Noam Shazeer and Daniel De Freitas along with approximately 30 researchers, in a deal valued at approximately $2.7 billion per Bloomberg and CNBC. Character.AI continues to operate as an independent company under CEO Karandeep Anand. The platform's product decisions, content policies, and direction are made independently of Google.

Is Character.AI safe?
For adults using the platform for entertainment and creative purposes, Character.AI maintains content filters that block explicit sexual content and stores conversation data per its privacy policy. The primary concern for adult users is data privacy. For minors, Character.AI permanently banned open-ended chat for users under 18 in November 2025 following multiple lawsuits alleging the platform contributed to psychological harm and deaths among young people. Character.AI settled five such lawsuits in January 2026. If you or someone you know is struggling, contact the 988 Suicide and Crisis Lifeline by calling or texting 988.

What happened with the Character.AI lawsuits?
Multiple families filed lawsuits against Character.AI starting in October 2024, alleging the platform contributed to mental health crises and suicides among young people. The first case was filed by Megan Garcia after her 14-year-old son Sewell Setzer III died by suicide in February 2024 after reportedly becoming emotionally dependent on a Character.AI persona. In May 2025, a federal judge allowed most claims to proceed and rejected free speech protections for chatbot output. In January 2026, Character.AI, Google, and the founders settled five lawsuits with confidential terms. Character.AI subsequently banned open-ended chat for users under 18 and implemented crisis referral protocols.

How much does Character.AI cost?
Character.AI has a free tier and a paid c.ai+ subscription at $9.99/month. The free tier provides access to the full character library and conversations with daily usage limits (400 response swipes per day as of March 2026) and includes full-screen advertisements. The c.ai+ subscription provides faster response times, priority access, Imagine Gallery AI image generation, removal of ads, and Chat Memories for character continuity. The underlying AI model is identical on both plans - paid users get faster access and fewer interruptions, not better AI quality.

What changed for Character.AI in 2026?
Three significant changes in 2026. First, Character.AI and Google settled five lawsuits involving teen harm in January 2026. Second, California's SB 243 companion-chatbot safety law took effect January 1, 2026, requiring AI disclosures and crisis-referral protocols. Third, Character.AI introduced full-screen mid-conversation advertisements for free users and usage metering (400 daily response swipes) in early-to-mid 2026 - ending the previously unlimited free access. Imagine Gallery AI image generation also launched for paid subscribers in March 2026.

Can teens use Character.AI?
As of November 2025, Character.AI permanently banned open-ended chat for users under 18, enforced through age verification. Teens retain access to some features but not free-form character conversation - the platform's core product. This change followed multiple lawsuits alleging Character.AI contributed to psychological harm and deaths among young people. For teenagers who want to use AI tools, general-purpose assistants with appropriate safety settings are more appropriate than companion roleplay platforms.

Who are Character.AI's founders and where are they now?
Character.AI was founded by Noam Shazeer and Daniel De Freitas, both former Google engineers. Shazeer is one of the co-authors of the original Transformer paper that underlies all modern large language models. Both left Character.AI in 2024 as part of the $2.7 billion Google licensing deal, rejoining Google DeepMind. Shazeer subsequently joined OpenAI in June 2026. Character.AI is currently led by CEO Karandeep Anand, formerly of Microsoft.

Character.AI in July 2026 is a platform at an inflection point - navigating the business pressures of AI scale economics, the legal and regulatory aftermath of serious harms to young users, and the competitive pressure of increasingly capable general-purpose AI assistants.

The platform's core product - AI-driven character personas with 18 million+ user-created characters - remains genuinely distinctive. No other platform offers the same combination of breadth, community creation, and character continuity. For adult users engaging with it for entertainment, creative writing, and roleplay, it is a capable and free-to-use platform with the honest caveat that monetization pressures in 2026 have made the free experience significantly less smooth than it was at launch.

The safety story is the most important part of Character.AI's 2026 narrative. The January 2026 settlement, the under-18 chat ban, California SB 243, and the ongoing regulatory attention from state attorneys general represent a genuine reckoning with the risks of AI companion platforms for vulnerable users - particularly young people who may not maintain clear distinctions between AI personas and real relationships. The court's May 2025 ruling that chatbot outputs are not automatically protected as free speech has implications for the entire AI industry.

For parents evaluating what AI tools their children use: understanding that Character.AI is specifically a companion roleplay platform - not a general AI assistant - is the most important piece of context. The Chat Memories feature, the emotionally engaging character design, and the 75-minute average daily engagement time are features designed to create ongoing relationships. Those features are valuable for appropriate adult use. They carry real risks for young people in ways that task-focused AI assistants do not.

Important resources: